Connect with us

Tech & AI

Milk is more regulated than AI, Microsoft’s Brad Smith tells the BBC

Published

on

brad smith microsoft president web summit stage.avif

brad smith microsoft president web summit stage

Microsoft’s president, Brad Smith, has called for global safety rules for AI. They include a legal duty to tell victims when a rogue AI agent breaks into their systems. He set out his case on the BBC’s Big Boss Interview podcast with Will Bain, released on 11 October.

Smith told the BBC: “Right now, your washing machine is subject to more safety rules than AI. Every carton of milk that you see in the grocery store is subject to safety rules and inspection.”

Smith called AI both a “tool and a weapon”. He said AI is moving forward “faster than the speed of safety”. That, he said, is not good for the public or for the industry itself.

Three measures

The first measure is a “safety brake or emergency brake”, often called a kill switch. It would cover any AI system that controls autonomous devices or critical infrastructure. Smith said it should sit with the companies running the models and in the cloud infrastructure they run on.

The second is a legal requirement to write “incident reports”. If AI agents “break loose” and break into systems, victims should be told. That already happens after a data breach. Anthropic said this month that one of its models sent a fake murder tip to police during testing.

Advertisement

Smith called the third measure the biggest and most important. It is whether government evaluators should sign off frontier models before release. He said cartons of milk and passenger aircraft are already subject to inspection and certification.

“We know how to do this for other products. We should build on that knowledge and think about how to apply it to AI,” Smith said.

Why the UK

Smith said the UK is best placed to drive the rules because of its AI Security Institute. Former prime minister Rishi Sunak’s government launched it as the AI Safety Institute. He said it has the strongest technical capability in the world to test AI models.

“It’s the United Kingdom. They are the best,” Smith said.

In Washington, he added, people do not look at London the way they look at Brussels. He called that a strength.

Last month, the White House asked OpenAI and Anthropic to hold back models from UK testers. On Monday, the UK said its testers would again get pre-release access.

Advertisement

At odds with Trump

President Donald Trump has said the industry should regulate itself so US firms do not lose out to Chinese rivals. He has repeatedly said that “whoever wins AI, wins”.

At the end of September, Trump hosted the leaders of OpenAI, Anthropic, Nvidia, SpaceX, Meta and Google. They signed a “morally binding” document at the White House. Microsoft did not sign it. Smith said he applauds the statement and that Microsoft is “happy to sign it”.

“It’s a good step, it’s a necessary step… But I don’t think it’s sufficient any longer,” Smith said. “It’s the companies that are not as responsible that you need to worry about.”

Smith made the same comparison to milk and washing machines in an interview with ITV News on 8 October. He told ITV that governments should be able to order AI systems offline, as they can ground aircraft. Developers should be legally liable if they act negligently and their AI causes catastrophic harm, he said.

Microsoft’s chief executive, Satya Nadella, called for an emergency brake on AI models on Saturday.

Advertisement

In the BBC interview, Smith also answered Vice President JD Vance, who accused Microsoft of abusing the H-1B visa system more than any other US company. Smith said Vance’s staff “got the maths wrong”. He said about 80% of Microsoft’s 6,000 visa applications were for existing employees, and that the company hired 9,000 Americans last year.

Source link

Advertisement
Continue Reading
Click to comment

You must be logged in to post a comment Login

Leave a Reply

Tech & AI

Want Your New iPhone Duo Fast? Lock in Your Details Now

Published

on

7564af8b 7eb5 5fb1 b0ab 4034066da597 thumbnail

If you’re eager to get your hands on the highly anticipated iPhone Duo, the first foldable iPhone, here’s how to get a head start on ordering. Apple has opened a staging page for preorders on apple.com, where you can get ready to place a preorder for the viral phone.

Head over to Apple’s online storefront, and you’ll find that a new iPhone Duo configurator is now live. On that page, save your preferred phone color, storage, mobile carrier and payment method. Saving these details in advance ensures you can complete your purchase immediately when preorders open at 5 a.m. PT on Friday.

Apple unveiled the iPhone Duo, one of its first new products under the helm of new CEO John Ternus, during its Surprise and Shine event in Cupertino, California, on Friday. It’s the first foldable from Apple, although merely the latest in a wave of new bendy phones to hit the market.

The iPhone Duo opens to reveal a 7.6-inch display and folds down to a pocketable passport size. The shifting shape supports split-screen multitasking and Apple Pencil, and brings a host of new features and animations to Apple iOS 27.

Advertisement

Pricing starts at $1,999 for the 512GB model and tops out at $3,199 for the 2TB version — before carrier subsidies — making the iPhone Duo the most expensive iPhone ever released. Price aside, observers expect demand among early adopters to be high, so prepping your order now is a smart idea.

iPhone Duo First Look: Amazing Battery, Features and Zero Crease?!

Not quite sure that the new iPhone Duo is the right phone for you? I’m also on the fence. Check out CNET’s coverage leading up to the launch of the new foldable and iOS 27, as well as the more conventionally shaped iPhone 18 Pro, so you can make an informed decision.

Source link

Continue Reading

Tech & AI

‘There’s definitely been a return to less technologically driven products’: Bose president Raza Haider talks releasing $99 wired earbuds with no app or AI, and why they’re only here because his daughter thinks they’re cool

Published

on

e97ffVZedmQ4CorkdLr6Mb

  • Bose recently launched $99 / £99 ‘Noise Cancelling Wired Earbuds’
  • Also, Bose is ‘happy to take credit for making open-ear mass market’
  • Bose’s President of Premium Audio spoke to me about both developments

Last month was huge for Bose’s earbuds output, thanks to the release of not one but two successors to the excellent Bose Ultra Open, a delightfully emo collab on the company’s flagship cans and a set of wired, inexpensive earbuds that boast ANC, controlled by the inline remote.

So hang on, on the one hand we’ve got an evolution on a groundbreaking, clip-style, wire-free product Bose pioneered just over two years ago, (something President Raza Haider says he’s “happy to take credit for — making open-ear mass market”). And on the other hand we have… wired earbuds like it’s 2003?

Source link

Advertisement
Continue Reading

Tech & AI

What’s The Difference Between LG OLED And OLED Evo TVs?

Published

on

intro 1791394511




LG is one of the biggest names in televisions (and also makes a number of wild products you may not even know about). It makes such a wide range of TVs, in fact, that its lineup can become confusing to the uninitiated. Case in point: LG OLEDs versus the company’s LG OLED Evo line. Both share the primary advantages of OLED, including exceptionally deep blacks and what amounts to nearly infinite contrast, as well as its biggest drawbacks. OLED is also known for its wide viewing angles, and because individual pixels make their own light instead of relying on a backlight, there is very little bloom on bright objects that are set against dark backgrounds.

In LG’s parlance, Evo doesn’t refer to a fundamentally different display technology, but rather the name for OLED models with additional, model-dependent brightness and processing enhancements. Evo first appeared in 2021 with the G1 series. According to LG, its Evo panels use an extra green-emitting material to increase peak brightness, as well as a specific deuterium-based organic material made to withstand higher voltages. Modern versions also now include processing upgrades alongside model-dependent thermal and light-control features.

Advertisement

What OLED Evo adds

The main upgrade Evo buyers will notice is brightness, though you need to make sure you’re not ruining your viewing experience with the wrong settings. Extra brightness can not only add punch in well-lit rooms but can make HDR highlights, like the sun glinting off the crest of a wave or a billowing explosion, look more pronounced. Some of the newest Evo sets include what LG calls Hyper Radiant Color Tech, which leverages the company’s alpha 11 AI processor to enhance color, brightness, and clarity.

According to LG, some of the latest G6 models can reach levels up to 3.9 times brighter than conventional OLED. The panels also offer Perfect Black, meaning no loss of definition even when images of different brightness are side by side, and Perfect Color, verified by independent testers at UL Solutions.

For a dark home theater, a conventional LG OLED can already deliver most of what you want from OLED: great black levels, contrast, and light control down to the pixel level. The extra cost of entry for an Evo really only makes sense if your TV will have to battle ambient light, or if you really prioritize eye-popping HDR. Regardless, things like the exact model and feature set of a panel will also play a huge role, likely more significant than just the presence or absence of the Evo label.

Advertisement



Source link

Advertisement
Continue Reading

Tech & AI

AMD puts 32,768 CPU cores in a rack built for AI agents

Published

on

amd illuminated logo sign escalator lobby.avif

amd illuminated logo sign escalator lobby

AMD has published Helion, an open rack blueprint for running AI agents rather than training models, with a CPU-only version holding up to 32,768 cores. Partner systems are due in November, and the pitch is how many agents fit in each watt.

AMD has published a rack blueprint for running AI agents rather than training models. It is called Helion, and partner systems are due in November.

The design appeared at the Open Compute Project summit in San Jose, The Register reported. It comes in a CPU-only version for hyperscale operators and a mixed CPU and GPU version for enterprises.

The CPU-only rack tops out at 32,768 cores. That is 128 sockets of AMD’s 256-core Venice chip on Zen 6, in 400W dense or 600W balanced blades.

Advertisement

The mixed rack is thinner on graphics than it sounds. It holds at most 16 Instinct MI350P accelerators, in two eight-GPU servers that AMD says run independently, with no scale-up fabric.

So this is not a frontier machine. AMD’s Helios rack carries 72 of the newer MI455X parts and 31 terabytes of HBM4, and remains the product for training.

Helion’s most interesting part is barely described. AMD says the enterprise rack carries a token router that sends agents’ model calls either to frontier models in the cloud or to open-source models running in the rack.

That detail appears only in AMD’s own release. Dan McNamara, who runs compute and enterprise AI at AMD, said the point was to let partners make efficient use of available power.

Advertisement

Power is the whole pitch. Matt Kimball at Moor Insights and Strategy said in a field note that the job is packing as many agents as possible into every watt.

He named sovereign AI operators among them, which puts Europe in the frame. Denmark’s grid operator paused new connections in March, with about 60GW of projects queued against peak demand near 7GW.

The constraint is not only Danish. Brussels adopted a Cloud and AI Development Act in June, aiming to at least triple the bloc’s data centre capacity within seven years.

Operators say the same thing. In a survey of 300 data centre decision-makers in the UK, Ireland and the Nordics, power availability came out as the leading obstacle to delivering sovereign capacity.

Advertisement

Helion is less finished than it reads. AMD has published no pricing, plans to contribute the specification to OCP rather than having done so, and named no European system builder among its partners.

Europe is still buying. The bloc has opened bidding for up to seven gigafactories at about €30B, each meant to hold 100,000 chips from suppliers including AMD.

Source link

Advertisement
Continue Reading

Tech & AI

KRAFTON Expands Esports College Campus to 256 Locations

Published

on

image004 1 1

KRAFTON India has announced the start of the 2026–27 Esports College Campus Tour, which now covers 256 colleges in India. The tour began on October 10, 2023, at IIT Kharagpur and IIT Varanasi, and is scheduled to end in March 2027. The new edition doubles last year’s reach, giving more college students a chance to compete in esports.

KRAFTON India Doubles Campus Tour Reach With BGMI and Real Cricket

KRAFTON India’s previous College Campus Tour reached 128 colleges and engaged over 75,000 students across India. This year, the company will double its reach and organize BGMI and Real Cricket tournaments across 256 colleges. The tournaments will give more students a chance to compete and represent their colleges in organized esports events. KRAFTON India is partnering with Nodwin Gaming and Upthrust Esports to run the tournaments. The events will take place alongside college festivals, making it easier for students to participate. College festival committees will manage the events with support from the organizers, while KRAFTON India will provide the prize pool.

Karan Pathak, Associate Director – Esports at KRAFTON India, said, “Last year, we saw how much interest there is among college students for competitive gaming, with the Campus Tour reaching 128 colleges and engaging over 75,000 students. This year, we are taking that a step further by reaching 256 colleges across the country. We want to give more students the chance to experience competitive gaming at their own campuses, compete with other colleges, and, for the strongest teams, progress to the next level.”

Winners to Advance to the College Campus Championship

The Campus Tour will follow a structured competition format, giving winning teams a chance to advance beyond their college-level tournaments. Winners from different colleges will play in zonal matches before advancing to the Grand Finals. After that, the best teams will move on to the Campus College Championship, where 16 teams will compete. The team that wins gets the opportunity to move forward directly to the invited stage of KRAFTON India’s official competition in 2027. The Campus Tour will run from October 2026 to March 2027, giving college teams across India several stages to compete and prove their skills.

Advertisement

Source link

Continue Reading

Tech & AI

How To Boost Your GPU Performance Without Buying A New One

Published

on

intro 1791488993

Your old GPU might not be able to compete with a new card, but it can run better with some tweaks.

In this increasingly anti-consumer tech world that seems fixed on abandoning all else in favor of data centers, there’s rarely been a worse time to buy a new GPU. While many panic-bought PCs to avoid RAMaggedon, there’s an even bigger crowd of PC gamers who have been unable to pick up a new graphics card for years. 

Just because you’re unable to afford a shiny NVIDIA RTX 50-Series GPU doesn’t mean you can’t squeeze more performance out of your current card, though. Some graphics tweaks, supersampling advice and general GPU maintenance will help boost FPS performance in your favorite games.

Advertisement

Use supersampling and frame generation to boost FPS

In an era where graphical advancements are outpacing hardware evolution — hello, path tracing — it’s becoming vital to get the most out of your current GPU through any means necessary. As someone who is now smitten with high-end ray tracing (it looks incredible in the likes of Cyberpunk 2077 or Marvel’s Spider-Man 2), I’m all about embracing AI to help boost frame rates. And yes, I do need to give my soul a shower for admitting that.

Whether you’re interested in advanced ray tracing features or not, it’s clear that graphically demanding games now rely on various forms of supersampling to lock down buttery, console-beating frame rates. Be it AMD FSR, NVIDIA DLSS or Intel XeSS, making the most out of AI-driven, FPS-boosting tech can be crucial for maintaining stable performance at high resolutions. None of us can be snobs when it comes to supersampling or frame generation now.

Both NVIDIA DLSS 4.5 and AMD FSR 4 produce image quality that’s on par with playing games at native resolution, while drastically boosting performance. These technologies work by rendering a game at a lower internal resolution (thus producing fewer pixels) before using deep learning techniques to upscale the final image to a far higher pixel count. Since DLSS 2.0 launched back in 2020, I have leaned on supersampling. With the current iterations of the big three solutions, it basically means free frames with little downside.

Advertisement

The issue of NVIDIA’s DLSS 4.5 Multi Frame Generation tech is a little more complicated. While it’s capable of producing 200 fps-topping results in Hogwarts Legacy with the x6 modifier enabled in my experience, it adds serious input latency in games that support the fledgling technology. Earlier incarnations of Team Green’s tech that top out at x2 and x3 modifiers perform nicely on supported titles, with the added benefit of producing fewer distracting visual artifacts than MFG (like shimmering around in-game text). 

As a PC player of 30 years, I can recommend supersampling without hesitation. Regarding the pros and cons of frame gen, I suggest experimenting with it until you make up your own mind.

Advertisement

Lower settings and turn off ray tracing

This is an obvious one, but if your GPU is sputtering away and dropping frames in the best PC games, it’s best to lower graphical settings right away. Only the fastest graphics cards can cope with a modern AAA title’s Ultra preset, so unless you’re rocking a top-tier card, it’s wise to drop your settings to Balanced. If your GPU is a couple of generations old, selecting a game’s Performance preset should help you claw back even more frames at the cost of graphical fidelity.

Unless you’ve got an ultra-obsessive eye, it’s unlikely you’re going to miss features like ambient occlusion or screen space reflections. By all means, tinker with in-game sliders to see what you like the look of and whether you can stomach the resulting performance cost.

In high-octane first-person shooters or fast-paced racing games, fancy rendering techniques are wasted on most gamers. Do yourself a favor and disable any setting that doesn’t enhance your experience. An increasing number of PC titles (like Resident Evil Requiem) have excellent graphics menus that provide before and after shots of how effects alter a game. If you can’t appreciate the difference in these preview panes, shut the setting off to improve your chances of boosting fps.

Those on budget GPUs should never enable ray tracing. This ultra-advanced lighting technique uses multi-bounce illumination to create incredibly realistic in-game shadows and reflections, but it’s hugely punishing on performance. Unless you own a beastly graphics card, ray tracing should almost always be disabled, as it can often lower frame rates by some 20 to 40 percent.

Advertisement

Remember the basics, clean your GPU and consider overclocking

Though the above methods should help pump up your frame rate numbers, there’s a physical element to this as well. You must know how to clean and organize your PC properly to get the most out of it. Keeping the inside of your system tidy can be a surprising way to up its effectiveness.

Thermal throttling is a potentially big issue for both your GPU and CPU. If you don’t regularly maintain your machine, dust can quickly build up, leading to a situation where the pride and joy of your gaming PC can’t maintain a regular airflow. Dust gathering around the fans of a GPU can lead to decreased performance as your card hits its temperature limit more quickly. Once it does, core clock speeds will reduce in order to lower temps, resulting in reduced frame rates.

To safely clean your graphics card, first switch off your PC and unplug it. Once you gain access to your GPU, you’ll want a can of compressed air and a soft microfiber cloth on hand. Use short spurts of air directed at the fans to disperse debris, then give the card a gentle wipe with the cloth. This isn’t a guaranteed FPS booster, but it will at least keep your GPU in good (hopefully optimal) shape going forward.

Don’t neglect the basics, either. Keep your GPU driver updated for the best performance, and close as many background apps as possible so your system isn’t devoting significant resources to anything outside your game. If you want to get advanced, overclocking your GPU can wring out more performance by going beyond the manufacturer’s intended settings. MSI Afterburner is the recommended tool for the job, though you’ll need to research how to do so safely.

Advertisement

Source link

Advertisement
Continue Reading

Tech & AI

Fervo’s Cape Station Geothermal Plant Goes Live

Published

on

a tall thin scaffold surrounded by tanks protrudes from the desert

This article was originally published by Canary Media.

The geothermal startup Fervo Energy is now selling power to the grid from its Cape Station project in Utah — marking the first time a utility-scale, next-generation geothermal facility has achieved such a feat worldwide.

The Houston-based firm turned on an initial 33-megawatt unit late last week and has been testing and ramping up output since then. On Thursday, Fervo will officially begin commercial operations, supplying carbon-free electricity to Shell Energy, a retail power provider in California.

Cape Station, a $2 billion project that will eventually be 500 MW, is unique among the world’s commercial geothermal plants. While other facilities draw on naturally occurring hot water and steam, Fervo has engineered its own geothermal resource deep underground. The startup uses drilling techniques from the oil and gas industry to create reservoirs in hot, dry rocks that generate steam to produce electricity.

Advertisement

“Traditional geothermal would have never been done in southern Utah,” says Dawn Owens, Fervo’s senior vice president and head of development and commercial markets.

With the first unit online, Fervo says it is working to commission two more 33-MW power blocks by 1 January, 2027. The project’s remaining 400 MW — which is supported by a new agreement with Google — is under construction and slated to come online in 2028.

Fervo’s Next-Generation Geothermal Journey

Fervo already operates a 3.5-MW commercial pilot plant in Nevada, which went online in late 2023. Owens says the pilot, called Project Red, was important for demonstrating the technology itself. But firing up Cape Station proves that Fervo’s approach can achieve utility scale and export power to wider markets.

“This is massive for both Fervo and the geothermal industry at large,” she says.

Advertisement

Fervo, which went public in May, is leading the fast-growing movement to develop clean, always-on geothermal power to meet soaring demand from data centers and balance growing supplies of intermittent solar and wind energy. Next-generation geothermal firms have raised billions of dollars in private capital in recent years to develop and deploy technologies that can harness energy in harsh, deep environments.

A tall, thin scaffold surrounded by tanks protrudes from the desert  Drilling continues at Fervo Energy’s Cape Station geothermal plant in Beaver County, Utah, where the company plans to ramp up production to 500 MW. Fervo Energy

Today, geothermal represents just 0.4 percent of total U.S. electricity generation every year. Expanding geothermal’s share of the mix will require not just building more projects like Cape Station but also tackling high upfront drilling costs and other economic and regulatory barriers — all while working to limit potential impacts on communities and the environment, experts say.

“Fervo’s first power achievement shows investors the size and scale of the opportunity to fuel many more such milestones across developers and sites in the future,” says Mitchell Smith, who serves as grid lead for the Mountain West Geothermal Consortium. The bipartisan initiative is calling for increased public and private financing “to help bring more projects online responsibly and at the speed of demand,” he added.

Enhanced Geothermal Energy Gets Cheaper

Fervo has declined to disclose the terms of its power purchase agreements for in-progress and future development, which together represent about 1,000 MW, or enough to power roughly 1 million U.S. homes every year. Phase 1 of Cape Station is set to deliver power at $7,000 per kilowatt, while Phase 2 should come at $5,500 per kilowatt, the company previously told Canary Media. (By comparison, the average cost to build a natural gas–fired power plant was $2,157 per kilowatt in 2025.)

Advertisement

Along with slashing drilling costs, enhanced geothermal projects could improve their economics by delivering power when it’s most valuable to utilities, according to researchers. While geothermal plants traditionally provide around-the-clock, baseload power, next-generation facilities might operate flexibly during certain periods — like in the evening, when solar output declines — to reduce overall grid costs. Geothermal plants could also operate “behind the meter” to directly power data centers, so that the costs of adding new electricity capacity aren’t passed on to utility customers, Owens says.

“We’ve been saying all along that our technology makes geothermal more scalable and affordable in more places,” she says. On Cape Station’s powering-up, she adds, “That’s just something no next-generation geothermal company has ever been able to do before.”

From Your Site Articles

Related Articles Around the Web

Advertisement

Source link

Continue Reading

Tech & AI

Don’t Buy A Cheap Android Phone Without These Non-Negotiable Features

Published

on

intro 1791485064

What good is a phone that can’t keep up with the average person’s day-to-day needs?

You don’t have to spend an arm and a leg on an Android phone, even if it’s brand-new. You can find plenty of great cheap Android phone options for $300 or less, if you know what to look for. And that’s the biggest thing to consider: knowing what features are crucial to have and where you can make concessions. Paying less for a phone that doesn’t have a sleek titanium frame, for example, is totally fine; you probably will (and should) put a case on it, anyway. But getting a phone with insufficient RAM, horrible battery life, poor cameras, a bad processor, or a small screen is where you want to draw the line.

In some cases, what you’re willing to sacrifice is an individual decision. If you’re a casual content creator, for example, the cameras need to be at least above average. But if you’re getting a secondary phone for basic messaging, you may be fine with just decent cameras. Nonetheless, there are some non-negotiable limits that anyone seeking out a cheap Android phone should focus on, no matter how you plan to use it.

Advertisement

Screen and RAM minimums

Remember when phones had tiny screens? The original iPhone’s 4.5-inch screen seemed groundbreaking at the time, but by today’s standards, that just doesn’t cut it. A phone should have at least a 5.6-inch screen, though most nowadays are at least 6.1 inches. This allows you to enjoy a comfortable experience when holding the phone and typing two-handed while still being able to clearly see the content on it. That screen should also be an AMOLED panel with HD resolution and a decent peak brightness of at least 1,000 nits, so it’s viewable in direct sunlight.

Memory is another factor to consider. You’ll want at least 6GB RAM; otherwise, multitasking, video streaming, and casual gaming could be slow or inconsistent, and the phone won’t respond quickly enough to your demands. If you play a lot of mobile games, consider even going up to 8GB. Also, if there’s no expandable storage (which is a rarity nowadays), 128GB is the minimum you should consider. For peace-of-mind, go with at least 256GB. Otherwise, you’ll end up with a “phone full” message before you know it, and you’ll be forced to clear out the device or sign up for cloud storage.

Finally, the processor matters. While you might not need the best, fastest, most up-to-date processor, you don’t want one that’s slow as molasses. Check the age of the processor, as well as reviews for its performance, and factor this into your decision. There’s no value in getting a cheap device if it’s slow and laggy when you want to stream a movie, and a great processor is vital for gaming on an Android phone.

Advertisement

Don’t skimp on the battery and cameras

Battery life and cameras are also important. A battery that’s at least 4,500mAh is sufficient, especially if the phone doesn’t handle heavy AI processing. But 5,000mAh is a better starting point nowadays to keep you going all day, even with moderate use. With cameras, you want a phone that can at least take photos that are good enough for sharing on social media and keeping for posterity. This means having at least a 32MP sensor for the rear camera, but also a large sensor that can capture more light and yield better photos overall, even if the megapixels are lower in one camera versus another. Don’t discount the front sensor and resolution as well, as this is what you’ll use to take selfies. Look for at least a 12MP front camera with, again, a good sensor size.

Consider other basics like NFC for tap-to-pay and more recent Bluetooth versions for reliable connectivity to headphones and speakers. For older models, confirm that there’s at least a few years left of software and security support. This sounds like a lot, but you’ll find plenty of devices that fit, like Samsung’s Galaxy A series phones (see our review of the Samsung Galaxy A17, for example) and phones from brands like Nothing and Motorola. 

Even older-model Google Pixel phones can fit the criteria. You can find a renewed Google Pixel 8a for just over $300 nowadays, and it comes with just about every vital feature mentioned previously. Ideally, wait for seasonal promotions to find a decked-out mid-range phone for the price of a budget one. You’ll have to make compromises to go cheap, but there are many phones in the $300 range that meet or exceed non-negotiable features and specs.

Advertisement

Source link

Continue Reading

Tech & AI

Zimacube 2 review: A tinkerer's NAS, at a good price point

Published

on

69238 146209 zimacube2review2

The ZimaCube 2 is a six-bay NAS with no software lock-in. At $799, the Standard model suits tinkerers, though the company’s Thunderbolt connection claims aren’t accurate.

Small silver cube-shaped computer case on a white desk, with front ventilation grilles and several ports including USB and audio jacks, against a plain wall background
ZimaCube 2 review

Once upon a time, I had a spare PowerMac 6500. Maybe it was a Performa — that specific is lost to me. This was just-pre MP3 explosion, but I still was generating a lot of data that I’d prefer not lurk constantly on my PowerBook 3400 first, then original PowerBook G3. What to do?
Then, in about the winter of 2002, hardware started aging less gracefully. The 6500 didn’t boot one day, so the project to move the home server to a Beige G3 accelerated. Then a series of G4 towers, and more hard drives than was thermally wise in a 1,1 Mac Pro.
Continue Reading on AppleInsider | Discuss on our Forums

Source link

Continue Reading

Tech & AI

This Anthropic Engineer Keeps Its AI Agent Trustworthy

Published

on

an indian man in a dress shirt smiling in front of the entrance to an office building

an indian man in a dress shirt smiling in front of the entrance to an office building

If you’ve ever wondered how Claude and similar AI agents can answer complex questions in seconds and where they get the information, you’re not alone.

Gurmeet Saran works to find that answer every day as an engineer at Anthropic in San Francisco, the AI research company behind Claude. The IEEE senior member is developing an infrastructure to understand the mechanics behind the chatbot’s behavior and to ensure its underlying architecture produces reliable outputs.

Gurmeet Saran

Employer

Advertisement

Anthropic in San Francisco

Title

Member of technical staff

Member grade

Advertisement

IEEE senior member

Alma maters

University of Kalyani, in India; Birla Institute of Technology and Science, in Pilani, India

“My job is to make sure the data is accurate, structured, and trustworthy, and that private data stays private,” he says. “As I learned earlier in my career, privacy can’t be a layer you bolt on—it has to be part of the architecture from the start.” If you can’t trust the logic, he adds, you can’t make reliable claims about the behavior, and that has real safety implications.

Advertisement

Safety is a priority for Anthropic, he says. According to Claude’s mission statement, the company wants the AI agent to be “genuinely helpful to the people it works with or on behalf of, as well as to society, while avoiding actions that are unsafe, unethical, or deceptive.” Crucial to the mission is ensuring that researchers can observe and understand how AI models behave, a challenge that is especially acute for large language models, Saran says.

“We have decent tools to monitor databases and web services,” he says, “but for LLMs, we’re still figuring it out.”

Solving data quality problems at Yahoo, Intuit, and Meta

Saran’s interest in computing systems dates back to his childhood in Kolkata, where he logged as much time as he could in the school computer lab.

“I remember the first time I used one of the computers,” he says. “What struck me was that you could write instructions, and the machine would follow them exactly. No ambiguity.

Advertisement

“That felt magical to me.”

After graduating from high school in 1999, he stayed in India to pursue advanced degrees. He earned a bachelor’s degree in electronics and telecommunication engineering in 2004 at the University of Kalyani. In 2010, he earned a master’s degree in software engineering from the Birla Institute of Technology and Science in Pilani.

While a graduate student, he began his career in 2007 at Yahoo as a technical lead of the data team in Bengaluru, India. While there, he encountered the project that shaped his career trajectory: working on web analytics and business intelligence reporting solutions.

“We kept running into a frustrating problem,” he recalls. “Different teams would calculate the same website metric—something like daily active site users—using slightly different logic, and end up with conflicting answers. It sounds like a small issue, but when leadership is making decisions about resource allocation and product development using that data, not knowing which number is correct becomes a problem.”

Advertisement

The team built a central data repository with defined metrics, essentially serving as a single source of truth. Saran created a model that defined what “daily active site users” means, breaking down which events counted, how to handle edge cases (or when standard code fails), and more.

With the new model, data teams across the company could rely on a standard method to determine the number of daily active users for the projects and sites they were responsible for and obtain comparable results using the same logic.

From there, Saran and his team built Yahoo! Digits, an internal web analytics tool used for all the company’s products.

Saran left Yahoo in 2010 to join Intuit as a staff engineer in Mountain View, Calif. He helped build the analytics infrastructure that unified data across the company’s products, including TurboTax and QuickBooks accounting software. During his tenure, Intuit developed a feature that detected fraudulent tax filings in real time. It required Saran and other engineers to build a system to protect taxpayer data while adhering to strict U.S. Internal Revenue Service policies.

Advertisement

“IEEE operates in a lot of countries where engineers don’t have easy pathways into the global professional community. That kind of connection matters as AI development spreads beyond a handful of organizations.”

“This taught me that privacy and governance can’t be layers you bolt onto software after the fact,” he says. “They have to be part of the architecture from the start.”

Saran joined Meta (known as Facebook at the time) in 2016 as a lead software engineer in Menlo Park, Calif. The company was facing controversy over its video metrics: Meta publicly acknowledged an error that caused it to overstate the number of views for posted videos.

“This wasn’t a coding bug; it was an architectural problem,” Saran says. He built a system that worked similarly to what he had developed at Intuit, whereby one single truth was defined to govern all video view metrics across the site.

Advertisement

That system, Metric360, eventually became part of the site’s architecture. By the time he left Meta in 2021, he says, Metric360 was being used across the company to calculate video metrics.

He left the company to become a principal software engineer at Wealthfront, a financial services company in Palo Alto, Calif. He helped build the company’s data infrastructure.

Building trustworthy AI systems at Anthropic

Saran joined Anthropic in 2025 as a member of the technical staff.

To describe the implications of his work, he uses a medical analogy: “Imagine a doctor is monitoring a patient’s vital signs and making treatment decisions based on those numbers. What if one or a few of the medical devices used were slightly miscalibrated? Even if most readings are fine, it’s a serious problem because the difficult decisions happen in the margins.

Advertisement

“It’s the same with AI systems. My job is essentially to make sure the instruments are giving accurate readings—that the data means what it says it means.”

His job comes with high stakes and complex problems to solve.

“AI agents are especially tricky to observe,” he says. “When a system is taking sequences of actions [generating conversational responses with different users], you need a way to capture what it did, what information it was working from, and whether outcomes matched what the end user expected. And you have to do all of that without exposing anyone’s private information—the systems are designed to recognize sensitive data and keep it protected as it flows through. It’s a much more demanding observability problem than anything I’ve worked on before.”

Solving for system reliability while AI is still in its early days of adoption is both Saran’s passion and the problem that keeps him up at night, he says.

Advertisement

“I think making AI systems genuinely auditable and transparent is one of the more important engineering problems of the next decade,” he says, “and it doesn’t get talked about as much as model capabilities do.”

IEEE can play a role in shaping AI development

Saran joined IEEE in March 2025.

“After more than 20 years of building data infrastructure across organizations like Yahoo, Intuit, Meta, and Wealthfront,” he says, “I wanted to contribute back to the technical community through peer review, publishing, and engaging with the broader engineering research ecosystem.”

He says he believes IEEE can play an essential role in shaping the future of AI through its standards work.

Advertisement

“We have standards for everything, from building codes to software testing,” he says. “AI infrastructure needs something similar, and IEEE has the credibility and the reach to do that.”

Another problem that IEEE can solve for, he says, is access: “IEEE operates in a lot of countries where engineers don’t have easy pathways into the global professional community. That kind of connection matters, especially as AI development spreads beyond a handful of organizations in a few cities.”

From Your Site Articles

Related Articles Around the Web

Advertisement

Source link

Continue Reading

Trending

Copyright © 2025